Printer Friendly

Currency substitution in Korea.

1. INTRODUCTION

The possibility of extensive currency substitution (CS) between domestic and foreign currencies has received a great deal of attention in the economics literature of industrialized nations in the past several years [see Bana and Handa (1987) for a review of these studies]. Very little empirical work, however, has been done in the case of developing countries. See, for example, Blejer (1978) and Ramirez-Rojas (1985), who found evidence in support of the CS hypothesis. The existence of CS has important implications for the working of flexible exchange rates. If the degree of CS is high, a small change in the money supply would induce large changes in the exchange rate. Furthermore, CS would transmit the effect of monetary disturbances from one nation to another.(1) Certainly, significant CS would seriously undermine the ability of flexible exchange rates to provide monetary independence.

This paper examines the empirical importance of CS in the framework of the demand function for money in the case of the developing economy of South Korea over the quarterly period 1973 through 1985. This is a useful exercise both for understanding the South Korean experience itself as well as for the examining the robustness of this proposition across countries with different economic and institutional frameworks.

Korea had a fixed exchange rate until 1979; during 1975 to 1979 the rate was pegged at W484 per U.S. dollar in spite of Korea's high domestic inflation compared to the rest of the world. On January 12, 1980, a managed floating system was adopted. In addition, a 20 percent devaluation of the won-to-U.S. dollar was undertaken. During the second quarter of 1980, as part of the Fifth Five-Year Plan, some reform of the country's financial sector was established. For example, accounts denominated in foreign currency were allowed and the total-surrender requirements of foreign receipts were relaxed for some agents. In addition, existing regulations on capital flows were further relaxed.(2)

The rest of the paper is organized as follows. Section 2 outlines the specification of the money demand model. The estimation results are presented and discussed in Setion 3. The conclusions are summarized in Section 4.

2. SPECIFICATION OF MONEY DEMAND EQUATION

A commonly used procedure to test for the importance of CS is to estimate a domestic demand for money and to determine if CS proxy significantly influences holdings of domestic real money balances. As noted by Bordo and Choudhri (1982), "If CS is important, the expected change in the exchange rate should be a significant determinant of the demand for home currency." The general estimating equation is of the form, (1) [Mathematical Expression Omitted]

where M = the domestic nominal
 money stock,
 P = the domestic price level,


(M/P) = the domestic real money
 stock,
 X = a measure of domestic real
 income,


[P.sup.e] = the expected rate of

inflation,

ER = a proxy for the expected
 change in the exchange rate
 --to capture CS,


[[(i.sup.d] - [i.sup.w]).sub.t] + [ER.sub.t] = a proxy for the expected
 nominal return on foreign
 bonds--to capture capital
 mobility,


[([i.sup.d] - [i.sup.w]).sub.t] = the domestic-foreign interest

rate differential, [d.sub.j] = seasonal dummy variables
 (J = 2,3,4),
 e = a disturbance term, and


In = the natural logarithm.(3)

The expected signs are [Mathematical Expression Omitted]

To estimate the above equation, some technical issues warrant comment. The first issue is how to define [P.sup.e]. In this study, [P.sup.e] is assumed to follow and adaptive expectations mechanism and is estimated using the Nugent-Glezakos (1979) method.

The second issue relates to how ER should be proxied. Several measures were tried. The first is the actual rate of devaluation of wons per U.S. dollar exchange rate. This is a rational expectations assumption that implies perfect foresight. The work of Edwards (1988:190) suggests that this assumption is not unreasonable for Korea.(4) The second proxy used in the once-lagged actual rate of devaluation of wons per U.S. dollar exchange rate.(5) The third is the current differential between inflation rates in South Korea and the United States measured in terms of wholesale price index [see Ramirez-Rojas (1982) for details]. The empirical results of using these proxies are shown in the appendix.

Finally, using Fair's (1987) recent procedure, the nominal adjustments specification was determined to be appropriate for the [M.sub.1] money demand. In addition, seasonal dummies were tried, but they were not important.

3. EMPIRICAL RESULTS

For space consideration, Table 1 presents the empirical results of estimating equation (1). The equation is preferred on the basis of [R[bar].sup.2], standard error of estimate (SEE), and diagnostic test results.(6)

As can be seen, the statistical fit of the equation to the data is excellent as indicated by value of [R.sup.2] (adjusted for degrees of freedom), SEE, and the F value for testing the null hyphothesis that all right-hand side variables, as a group, except the constant term, have a zero coefficient. According to the Durbin-Watson test, there is no evidence of significant first-order serial correlation in the residuals. This finding is supported by the Durbin-H statistic. As argued by Wallis (see Johnston 1984:317) where quarterly data is employed, it is important to test for the fourth-order autocorrection. The Wallis [d.sub.4] of 1.867 is not significant at the 5 percent level.

As evident in Table 1, the signs of the estimated coefficients on all explanatory variables (except [ER.sub.1t]) are consistent with a priori predictions of the theory. An encouraging aspect of our results is that the proxy for expected change in the exchange rate appears with the expected negative coefficient which is significantly different from zero at thee 5 percent level in the financial deregulation period. The results, therefore, indicate that there is some support for CS. The long-run elasticity [Mathematical Expression Omitted] multiplied by the ratio of the arithmetic means of the [ER.sub.2] to dependent variable] is -0.075. The estimate coefficient on [ER.sub.1] is not statistically significant.

Furthermore, note that real income, expected inflation, and the yield on interest-bearing foreign assets have the theoretically-expected sign and are statistically significant at the 5 percent level. The long-run real income elasticity is 0.34 and implies that money holdings are not viewed as a luxury type of good.(7) The long-run elasticities of expected inflation and yield on foreign-interest bearing assets are -0.025 and -0.02, respectively. [Note that these elasticities are calculated as those of [ER.sub.2] above.] The estimated speed of adjustment (1 - [[Psi].sub.6] is about 47 percent per quarter.

In applied econometric research, it is not uncommon to estimate a totally meaningless model and yet obtain "correct signs" and high coefficient of multiple determination. This is in line with Granger and Newbold (1974) and Lovell (1983) who point out the ease with which high t-values can be obtained without there being any relationships between variables whatsoever. For example, the high-adjusted coefficient of determination shown in Table 1 may be due to spurious correlation arising from the model being inappropriately specified. Kramer et al. (1985) have suggested that conventional regression output be supplemented with a battery of specification tests, since this will make it harder for results to appear significant due to a researcher's intentional or unintentional "data mining" process. [Tabular Data 1 Omitted]

In order to further assess the robustness of the results presented above, the models have been tested to check the constancy of the estimated coefficients, higher-order autocorrelation, heteroscedasticity, omitted variables, and the sensitivity to the dynamic specification. The results of these tests are presented in Table 2. [Tabular Data 2 Omitted]

As may be seen, none of the test statistics is significantly different from zero at the 5 percent level. In sum, they lend credence to the empirical findings of this study.

4. CONCLUSIONS

In this study, a modest attempt has been made to examine explicitly the effects of currency substitution on the money demand function. The results lend support to the proposition that, in the wake of financial deregulation, currency substitution undoubtedly has a negative effect on the demand for money in South Korea. In closing, it seems appropriate to refer to similar "significant" results from elsewhere in the world: Melvin (1985) in the Western European currency union context, Batten and Hafer (1984) for Canada and Germany, Browne (1986) in the Irish context, Ghosh (1989) in the Canadian context, and Blejer (1978) and Ramirez-Rojas (1985) in the Latin American context, to name only a few.

APPENDIX A

DEFINITION OF VARIABLES AND SOURCES OF DATA

The variables used in this study are all measured and defined empirically as follows:

[M.sub.1] = nominal money stock (line 34).
 P = consumer price index, 1980 = 100 (line 64).
 X = real income (line 99b.p).


[P.sup.e] = derived using quadratic loss function.
 The rate of inflation is measured as ([InP.sub.t]
 - [InP.sub.t-1])(*) 400.
 ER = the expected change in the exchange rate.
 Period averages (line rf) was used. In
 addition, the current differential between the
 inflation rates in Korea and the United
 States was used also. Both variables are
 measure in terms of wholesale price
 index (line 63).


[i.sup.w] = short-term world interest rates. This
 variable is constructed as the average of
 quarterly short-term interest rates of 9 countries
 which are the main trading partners of
 Korea (Canada, France, Germany, Japan,
 Netherlands, Sweden, Switzerland, the
 United Kingdom, and the United States).
 The rates are given in Madura (1989:
 612-61).


[i.sup.d] = domestic interest rate (line 60)

The line numbers indicated above refer to the (IFS) International Financial Statistics published monthly by International Monetary Fund. Monthly data on money stock were use to derive the middle-of-quarter figures since other flow variables are middle of quarter measures. Exchange rate change at time (t) is [Mathematical Expression Omitted] [Tabular Data B to D Omitted]

Notes

(1)See Batten and Hafer (1984) for details. (2)On the recent evolution of Korea's financial sector, see Edwards (1988, 1989); Cho (1985); and Economic Planning Board (1982). (3)For a similar estimating equation that allows for differential effects of CS between fixed and flexible exchange rate regimes, see Batten and Hafer (1984) and Bana and Handa (1987). (4)As Warner and Kreinin (1983:97) pointed out, because of the visibility of exchange rate movements, market participants may be more aware of them than they are of other price changes. (5)This measures assumes adaptive expectations with full adjustment to errors. The Nugent and Glezakos procedure gave a coefficient of adjustment of 0.95. (6)Mention should be made that the use of the actual rate of devaluation or its lagged value turned up to be significant and appropriately signed. However, these equations failed the Hendry-Pagan-Sargan test of dynamic specification. Using Japan as the reference country and modelling ER by the errors-in-variables method of rational expectations and lagged foreign interest rates yielded similar results. These results are contained in a longer version of this paper. (7)Permanent income was also used as the scale variable instead of current income. This variable was generated in two ways. First, a time-series forecast of X was used to produce a proxy for permanent income. An AR(2) was found appropriate (introducing further lags did not improve the fit of the question). Second, the Nugent-Glezakos (1979) was used. The value that minimized this quadratic loss function was 0.97. Reestimating the equations with permanent income proxies did not improve the fit of the equations.

References

Bana, I. M., and Handa, J. (1987) "Currency Substitution: A Multicurrency Study of Canada." International Economic Journal, Autumn, 71-86. Batten, Dallas, S., and Hafer, R. W. (1984) "Currency Substitution: A Test of Its Importance." Federal Reserve Bank of St. Louis, Review, August/September, 5-12. Blejer, Mario I. (1978) "Black Market, Exchange Rate Expectations and the Domestic Demand for Money." Journal of Monetary Economics, 4, 747-763. Bordo, M. D., and Choudhri, E. U. (1982) "Currency Substitution and the Demand for Money: Some Evidence for Canada." Journal of Money, Credit and Banking, 14 (February), 48-57. Browne, F. X. (1986) "Multilateral Currency Substitution and Capital Flows as Sources of Instability in the SOE Demand for Money Function--A Case Study." Empirical Economics, 11, 181-196. Cho, Y. J. (1985) "Capital Market Structure and Barriers to Financial Liberalization." Unpublished manuscript. Washington, DC: The World Bank. Edwards, S. (1989) Real Exchange Rates, Devaluation, and Adjustment. The MIT Press, Massachusetts. Edwards, S. (1988) "Financial Deregulation and Segmented Capital Markets: The Case of Korea." World Development, 16, 185-194. Fair, R. C. (1987) "International Evidence on the Demand for Money." Review of Economics and Statistics, 69 (August), 473-480. Farley, J. H., Hinich, M. J., and McGuire, T. W. (1975) "Some Comparisons of Tests for a Shift in the Slope of a Multivariate Linear Time Series Model." Journal of Econometrics, 3, 287-318. Ghosh, Sukesh, K. (1989) "Currency Substitution and Demand for Money in Canada: Further Evidence." Journal of Macroeconomics, II (Winter), 81-89. Granger, C. W. J., and Newbold, P. (1974) "Spurious Regressions in Econometrics." Journal of Econometrics, 2, 111-120. Hausman, J. A. (1978) "Specification Tests in Econometrics." Econometrica, 46, 327-349. Hendry, David F., Pagan, Adrian R., and Sargan, Denis J. (1984) "Dynamic Specification." In Handbook of Econometrics. Amsterdam: North-Holland Publishing Company. Johnson, J. (1984) Econometric Methods, Third Edition, New York: McGraw Hill. Kramer, W., Sonnberger, H., and Havlik, P. (1985) "Dynamic Checking in Practice." Review of Economics and Statistics, 67 (February), 118-123. Lovell, M. D. (1983) "Data Mining." Review of Economics and Statistics, 65 (February), 1-12. Madura, Jeff. (1989) International Financial Management, Second Edition. New York: West Publishing. Melvin, M. (1985) "Currency Substitution, and Western European Monetary Unification." Economica, 52 (February), 79-91. Nugent, J., and Glezakos, C. (1979) "A Model of Inflation and Expectation in Latin America." Journal of Development Economics, 6, 431-446. Ramirez-Rojas, C. L. (1985) "Currency Substitution in Argentina, Mexico, and Uruguay." IMF Staff Papers, 43 (December), 629-667. Ramsey, James B. (1969) "Tests for Specification Errors in Classical Linear Least Squares Regression Analysis." Journal of the Royal Statistical Society B, 31(2), 350-371. Sargan, J. D. (1958) "The Estimation of Economic Relationships Using Instrumental Variables." Econometrica, 36, 393-413. Warner, D., and Kreinin, M. (1983) "Determinants of International Trade Flows." Review of Economics and Statistics, 65 (February), 96-104. Wu, P. (1973) "Alternative Tests of Independence Between Stochastic Regressors and Disturbances." Econometrica, 41, 733-750. Augustine C. Arize College of Business and Technology East Texas State University. The author would like to thank the following for helpful comments on earlier drafts: Ed Manton, Ray Ballard, and J. B. Spalding. Special thanks to Trezzie Pressley for continued financial support and to Nancy Swanson for excellent typing. Bedford Umezulike provided able research assistance.
COPYRIGHT 1991 Omicron Delta Epsilon
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1991 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Arize, Augustine C.
Publication:American Economist
Date:Sep 22, 1991
Words:2453
Previous Article:The impact of imports on price competition in the automobile industry.
Next Article:Investment in name brand capital: evidence from the free banking era.
Topics:


Related Articles
Flexsys files $20 million patent infringement suit.

Terms of use | Copyright © 2016 Farlex, Inc. | Feedback | For webmasters